Structural, mechanical and piezoelectric properties of polycrystalline AlN films sputtered on titanium bottom electrodes

G Model

ARTICLE IN PRESS

APSUSC-30747; No. of Pages 12

Applied Surface Science xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

Applied Surface Science journal homepage: www.elsevier.com/locate/apsusc

Structural, mechanical and piezoelectric properties of polycrystalline AlN films sputtered on titanium bottom electrodes M. P˘atru a,∗ , L. Isac a , L. Cunha b , P. Martins b , S. Lanceros-Mendez b , G. Oncioiu c , D. Cristea a , D. Munteanu a a

Department of Materials Science, “Transilvania” University, 29 Eroilor blvd., 500036 Brasov, Romania Centre/Department of Physics, University of Minho, Campus de Gualtar, 4710-057 Braga, Portugal c Institute for Nuclear Research Pitesti, 1 Campului Str., 115400 Mioveni, Romania b

a r t i c l e

i n f o

Article history: Received 23 December 2014 Received in revised form 3 July 2015 Accepted 4 July 2015 Available online xxx Keywords: AlN films Piezoelectricity Nanoindentation Thermal stability

a b s t r a c t Polycrystalline AlN coatings were deposited on Ti-electrode films by reactive magnetron sputtering. During the deposition, processing parameters such as the reactive gas pressure and time of deposition have been varied. The purpose was to obtain an optimized AlN/Ti system coating with suitable properties for applications such as piezoelectric sensors, which could monitor the wear rate and the remaining coating life of a specific part. The chemical composition, the structure, and the morphology of the multilayered films were investigated by X-ray photoelectron spectroscopy, X-ray diffraction, scanning electron microscopy and atomic force microscopy techniques, respectively. These measurements showed the formation of highly (1 0 1), (1 0 2) and (1 0 3) oriented AlN films with piezoelectric and mechanical properties suitable for the desired purpose. A densification of the AlN coating was also observed, caused by lower nitrogen pressures, which has led to an improvement of the crystallinity along with an increase of hardness. The coating stability at high temperatures was also examined. Consequently, an improvement of the piezoelectric properties of the AlN films was observed, inferred from the enhancement of c-axis (0 0 2) orientation after annealing. Furthermore, the mechanical characteristics (hardness and Young’s modulus) were significantly improved after heat treatment. These two parameters decrease rapidly with the increase of the indentation depth, approaching constant values close to those of the substrate after annealing. Thus, thermal annealing promotes not only the rearrangement of Al–N network, but also a surface hardening of the film, caused by a nitriding process of unsaturated Al atoms. © 2015 Elsevier B.V. All rights reserved.

1. Introduction The wear mechanism of a tribological system is very complex. There are two broad approaches concerning the study of wear: the first is based on describing the effect on tribological performance of various parameters like counterpart, coating (if any), substrate, application conditions and environment, while the second is based on the physical nature of the underlying processes. The solution to wear related problems should start with a detailed examination of the tribological system with all the factors that are involved. Even if these factors are not constant, possibly leading to the premature failure of the system, it is very important for a designer to know the amount a component can be worn before it must be

∗ Corresponding author. E-mail address: [email protected] (M. P˘atru).

replaced. In the case of coated parts, wear monitoring of the coating could provide the coating status life and also a warning before the coating failure occurs, thus conferring a higher degree of reliability to the tribological system. Embedding a piezoelectric layer between two electrodes into a tribological coating system to monitor the stress variation within the material could be an appealing technique for in situ wear monitoring and, consequently, would potentially help to develop new materials by providing additional information about wear mechanisms. Aluminum nitride AlN is used in many applications, in microelectronic and optoelectronic devices such as ultraviolet detectors, light emitting diodes [1], thermal interface materials [2] and piezoelectric materials in surface acoustic wave devices [3]. Several techniques like molecular beam epitaxy (MBE), chemical vapor deposition (CVD), pulsed laser deposition (PLD) and reactive sputtering [4–7] can be used to deposit AlN thin films. This type of compound is a promising candidate as a piezoelectric material, which, considering its other remarkable properties (large optical

http://dx.doi.org/10.1016/j.apsusc.2015.07.022 0169-4332/© 2015 Elsevier B.V. All rights reserved.

Please cite this article in press as: M. P˘atru, et al., Structural, mechanical and piezoelectric properties of polycrystalline AlN films sputtered on titanium bottom electrodes, Appl. Surf. Sci. (2015), http://dx.doi.org/10.1016/j.apsusc.2015.07.022

G Model

ARTICLE IN PRESS

APSUSC-30747; No. of Pages 12

M. P˘atru et al. / Applied Surface Science xxx (2015) xxx–xxx

2 Table 1 Deposition parameters of Ti/AlN coating-systems.

Ti electrode films

Target material Sputtering power Working pressure Gas pressure Deposition rate Sputtering time Substrate

Ti – 99.99% purity 1.2 kW 5 × 10−4 mbar Ar – 0.25 mbar 0.8 ␮m/h 1h Steel 100Cr6

AlN piezoelectric films Series no.1

Series no. 2

Series no. 3

Al – 99.99% purity 1.5 kW 5 × 10−4 mbar N2 – 0.1 mbar 1.2 ␮m/h 2h Ti-film

Al – 99.99% purity 1.5 kW 5 × 10−4 mbar N2 – 0.05 mbar 1.2 ␮m/h 2h Ti-film

Al – 99.99% purity 1.5 kW 5 × 10−4 mbar N2 – 0.1 mbar 1.2 ␮m/h 4h Ti-film

band gap, high hardness and good thermal conductivity), should be successfully used in the fabrication of a piezoelectric wear monitoring sensor. The deposition by sputtering of AlN films with controlled crystal orientation is one of the key issues intensely studied, considering the fact that the piezoelectric properties of such films are strongly dependent on the crystallographic structure. The influence of the deposition parameters such as target power, growth temperature, sputtering pressure and gas composition on the properties of aluminum nitride films has been reported, to some extent, for films deposited by DC [8,9] and RF reactive sputtering [10–14]. According to the literature it seems that the physical mechanism that determines the way in which the Al and N atoms are structurally arranged is related to the kinetic energy of the adatoms during the film growth [15]. At high kinetic energy the AlN films grow with the c-axis normal to the surface ((0 0 2) orientation). As the energy is decreased, the c-axis of the crystals gradually tilts away from the normal direction, resulting in other preferred orientations such as (1 0 3), (1 0 2), (1 0 1), and, eventually, (1 0 0) or (1 1 0) orientations [15]. Nevertheless, a convincing correlation between the sputtering parameters and the crystal orientation of sputtered AlN films, to the best of our knowledge, has not been reported yet. Furthermore, contrasting results can be found in the literature. For example, if the sputtering pressure is considered, Xu et al. [16] reported that the (0 0 2) texture is improved at low sputtering pressures, while Cheng et al. [17] suggested the use of a moderate pressure for the same purpose. If we consider the nitrogen flow, Cheng et al. [17] report the deposition of AlN films with high (0 0 2) orientation when using high nitrogen flows, while Okano et al. [18] reported that the c-axis orientation is improved by decreasing the nitrogen concentration in order to avoid the formation of AlN on the aluminum target (target poisoning). The majority of previous works concluded that the AlN films must exhibit high (0 0 2) orientation in order to obtain the best piezoelectric properties. In this context, the purpose of the present work is to demonstrate that AlN polycrystalline films produced by DC reactive magnetron sputtering exhibit good piezoelectric properties when a high percentage of coating crystallinity is observed, despite its crystalline orientation. In addition to its piezoelectric properties, the successful fabrication of devices based on AlN thin films requires a better understanding of how the microstructure of the deposited material relates to the mechanical behavior of the film, which determines the practical use of a material. Moreover, if this piezoelectric device based on AlN/Ti films will be integrated into wear resistant coatings, then the as-deposited AlN film has to go through another deposition process to build up the tribological coating on top of it. Thus, the properties of AlN films might not only be influenced by the deposition parameters but also could depend on post deposition heat treatments, such as annealing. In this paper we will present our results concerning the changes in the structural, mechanical and electrical behavior of magnetron sputtered AlN films, subjected to heat treatments.

A series of AlN coatings were deposited using different sputtering conditions on Ti electrode films, in order to obtain an optimized AlN/Ti system coating, suitable for integration into wear resistant systems. Titanium has been selected as bottom electrode because it is a complementary metal–oxide–semiconductor (CMOS) compatible with AlN. Titanium exhibits in the crystalline state a lattice mismatch of 5% to AlN, as opposed to 23% in the case of pure aluminum, thus reducing the mechanical stress near the film interface [19]. The applicability of titanium as bottom electrode is also due to its excellent adhesion properties [21]. The feasibility of using sputtered Ti thin films as bottom electrode was reported elsewhere. Ababneh et al. [20] conclude that the Ti layer surface morphology variation, dependent on different deposition parameters, has a significant influence on the piezoelectric coefficients exhibited by the top AlN coating, synthesized under fixed sputter deposition conditions. Consequently, for this paper, the Ti underlayer was prepared using fixed sputtering parameters, limiting its influence on the AlN coating. This factor allowed us to correlate the change of structure and physical properties, shown by the AlN films, to the deposition parameters and subsequent post deposition heat treatments. 20 4400

25

30

35

40

45

50

55

60

65

70

Coating no.3 N2 pressure: 0.1 mbar Sputtering time: 4h

3300 2200 1100

Intensity (abs. units)

Deposition parameters

Coating no.2 N2 pressure: 0.05 mbar Sputtering time: 2h

4200 2800 1400 0

Coating no.1 N2 pressure: 0.1 mbar Sputtering time: 2h

4000 3000

Ti (200)

AlN(103)

AlN(102)

AlN(101) AlN(002)

2000 1000 20

25

30

35

40

45

50

55

60

65

70

Angle 2θ (deg.) Fig. 1. XRD spectra of the as deposited AlN films.

Please cite this article in press as: M. P˘atru, et al., Structural, mechanical and piezoelectric properties of polycrystalline AlN films sputtered on titanium bottom electrodes, Appl. Surf. Sci. (2015), http://dx.doi.org/10.1016/j.apsusc.2015.07.022

G Model

ARTICLE IN PRESS

APSUSC-30747; No. of Pages 12

M. P˘atru et al. / Applied Surface Science xxx (2015) xxx–xxx

3

Table 2 XRD analysis results. Crystallinity As deposited

No. 1

54.8%

Crystal phase composition Annealed at 900 ◦ C

61.8%

No. 2

56.1%

63.3%

No. 3

37.2%

54.7%

As deposited

36.64% AlN 63.36% Ti

32.15% AlN 67.85% Ti

68.15% AlN 31.85% Ti

47.16% AlN 50.84% Ti

33.31% AlN 66.69% Ti

AlN orientation

Annealed at 900 ◦ C

27.70% AlN 72.30% Ti

2. Experimental details Three series of AlN coatings were deposited on Ti electrode films by DC reactive magnetron sputtering. The variable processing parameters were the reactive gas pressure and time of deposition. The base support for the AlN/Ti multi-layer systems was 100Cr6 steel. After the deposition of the titanium electrode, with a thickness of about 300 nm, the aluminum nitride films were synthesized from a high purity Al target, at the working pressure of 5 × 10−4 mbar. Nitrogen was used as both the reactive and the sputtering gas. The sputtering power was set at 1.5 kW and the deposition rate was kept at 1.2 ␮m/h for each prepared sample. The nitrogen pressure and sputtering time were varied according to Table 1, in order to investigate their influence on crystallinity, morphology, mechanical and piezoelectric properties of the aluminum nitride films. The Ti thin films were synthesized under fixed sputter deposition conditions for all three series of AlN coatings. The deposition parameters are summarized in Table 1. The structure of the AlN films was investigated by X-ray diffraction (XRD, Bruker D8 Discover Diffractometer that uses Cu-K␣1 radiation). The crystallinity and crystal phase composition (in %) of the films was computed from XRD scans using DIFFRAC.EVA.V 1.4 software. X-ray photoelectron spectroscopy (XPS – ESCALAB 250 spectrometer with monochromatic Al K␣ radiation and 0.1 eV passing energy) was used to determine the chemical composition of the AlN films as well as aluminum and nitrogen atom bonding. As any contamination could produce deviation from the real chemical composition, the XPS analysis was performed in ultra-high vacuum conditions with a sputter cleaning source to remove any undesired contaminants. The surface morphology and roughness of the samples were studied using a scanning electron microscope (SEM) equipped with energy dispersive X-Ray (EDX) spectroscopy (SEM model FEI Inspect S) and atomic force microscopy (AFM NT-MDT model BL222RNTE). The AFM images were taken in semi contact mode, with the following protocol: Si-tips (NSG10, force constant 0.15 N/m, tip radius 10 nm). The results were processed with NTMDT Nova Soft software. The mechanical properties of the AlN coatings were measured using an Agilent G200 Nanoindenter, equipped with the CSM (Continuous Stiffness Measurement) module and with a Berkovich diamond tip (˛ = 65.3 ± 0.3◦ ). The continuous contact stiffness measurement (CSM) procedure [21] was followed, which was accomplished by superimposing small oscillations at 45 Hz on the force signal to measure displacement responses. The indenter was loaded with a strain rate of 0.05 s−1 . The piezoelectric response measurements were performed with an APC International d33 meter at room temperature (25 ◦ C), on

Annealed at 900 ◦ C

As deposited (h k l)

FWHM (◦ )

D (nm)

(1 0 1) (1 0 2) (1 0 3) (0 0 2) (1 0 1) (1 0 2) (1 0 3) (0 0 2)

0.685 0.311 0.761 0.970 0.550 0.360 0.731 0.919

10.96 23.18 8.77 7.80 13.65 20.03 9.13 8.24

(1 0 2) (1 0 3) (0 0 2)

0.303 0.993 1.087

23.79 6.72 6.96

(h k l)

FWHM (◦ )

D (nm)

(1 0 2) (1 0 3) (0 0 2)

0.338 0.443 0.511

21.33 15.05 14.78

(1 0 3) (0 0 2)

0.621 0.406

10.74 18.61

(1 0 2) (1 0 3) (0 0 2)

0.394 0.582 0.514

18.30 11.46 14.69

the as-deposited and annealed samples. Each post-deposition thermal annealing on the AlN films was carried out in a Thermolab vacuum furnace, under a base pressure of 10−6 mbar, with the following protocol: (a) increase of the furnace temperature, at a rate of 5 ◦ C/min, up to the desired temperature (300 ◦ C, 600 ◦ C and 900 ◦ C); (b) 1 h annealing at the desired temperature; and (c) free cooling down to the room temperature. 3. Results and discussion 3.1. XRD analysis As a general overview, the structural characterization reveals that all AlN films deposited on the Ti layer have a hexagonal AlN

20 4400

25

30

35

40

45

50

55

60

65

70

Coating no.3 N2 pressure: 0.1 mbar Sputtering time: 4h

3300 2200 1100

Intensity (abs. units)

Series of AlN coatings

Coating no.2 N2 pressure: 0.05 mbar Sputtering time: 2h

4200 2800 1400 0

Coating no.1 N2 pressure: 0.1 mbar Sputtering time: 2h

4000

Ti (200)

3000 Ti (220) AlN(002)

2000

AlN(102)

AlN(103)

AlN(101)

1000 20

25

30

35

40

45

50

55

60

65

70

Angle 2θ (deg.) Fig. 2. XRD spectra of the AlN films after annealing at 900 ◦ C.

Please cite this article in press as: M. P˘atru, et al., Structural, mechanical and piezoelectric properties of polycrystalline AlN films sputtered on titanium bottom electrodes, Appl. Surf. Sci. (2015), http://dx.doi.org/10.1016/j.apsusc.2015.07.022

G Model APSUSC-30747; No. of Pages 12

ARTICLE IN PRESS M. P˘atru et al. / Applied Surface Science xxx (2015) xxx–xxx

4 Table 3 XPS analysis results. Deposition condition of AlN coatings

Chemical composition of the AlN coatings Al 2p (%)

N 1s (%)

C 1s (%)

O 1s (%)

No. 1 N2 pressure: 0.1 mbar Sputtering time: 2 h

As-deposited Annealed at 900 ◦ C

41.07 39.98

26.43 28.58

9.40 10.56

23.10 20.88

No. 2N2 pressure: 0.05 mbarSputtering time: 2 h

As-deposited Annealed at 900 ◦ C

46.24 37.56

33.68 37.99

6.75 8.65

14.33 15.80

No. 3 N2 pressure: 0.1 mbar Sputtering time: 4 h

As-deposited Annealed at 900 ◦ C

40.37 36.30

27.91 29.52

7.29 9.47

24.43 24.71

structure. No other crystalline phase was found in the measured range, with the exception of cubic Ti, due to the presence of the bottom electrode film. From the XRD patterns, shown in Fig. 1, one can notice that the growth of the AlN films is preferentially orientated along the (1 0 1), (1 0 2) and (1 0 3) planes. A weak c-axis orientation growth along (0 0 2) plane is also registered. Furthermore, several other observations can be made regarding the structural development. The as-deposited films obtained with different nitrogen pressures (samples 1 and 2) show differences related to the intensity and the full-width at half-maximum (FWHM) of the peak signals. From Table 2, one can observe that reducing the nitrogen pressure from 0.1 to 0.05 mbar leads to an increase in the intensity of the AlN peaks and a decrease of the (FWHM). This observation suggests that an improvement in the crystal quality of the AlN coatings is occurring. These differences could be attributed to the variations of the amounts of energies created and stored in the coatings at different nitrogen deposition pressures, resulting in various phase developments and densification of the coating structures. As the reactive gas pressure increases, the sputtered atoms become more heavily scattered and arrive at the substrate surface with reduced energy. Consequently, the coating nucleation and crystalline quality suffer. This might be the reason why the coatings deposited at lower nitrogen pressure show the highest degree of crystal orientation, but also the highest content of AlN in the crystal phases composition (coating no. 2 – Table 2). If the deposition time is increased, from 2 to 4 h, a slight increase of the FWHM is observed, which is associated with a lower crystallinity of the film. This structural behavior can be explained by the possibility that during a longer sputtering time the fine crystallites grown at the beginning of the sputtering process start to aggregate into thicker grains, enhancing also the growth of amorphous material in the grain boundaries. The size of the AlN grains was also calculated using the Scherrer’s equation [22]. Therefore, the grain size evolution seems to be influenced by the deposition conditions (nitrogen pressure and sputtering time) and may act as an important controlling parameter. The crystalline film quality was determined by the intensity and the FWHM, considering the fact that the most significant sign for achieving good piezoelectric coefficients is a high degree of crystallinity of the coating. Akiyama et al. [23] reported also that a high degree of crystal orientation in a piezoelectric material results in a good electromechanical coupling coefficient. The post-deposition annealing at 900 ◦ C induced an overall improvement in the crystal quality of all AlN films. The FWHM of the (0 0 2) rocking curve for the annealed films is around 0.5◦ . This value is smaller than the values recently reported for AlN deposited on Ti layers [24,25]. Analyzing the XRD patterns of the samples after annealing, two different behaviors are observed which leads to the classification of the analyzed coatings in two categories. Films of type I are those that exhibit in the as-deposited state a high crystallinity above 50%. On the other hand, films of type II show a lower crystallinity, less than 50%. The type I polycrystalline AlN films, which are highly oriented along (1 0 1), (1 0 2) and (1 0 3) planes in the as-deposited state,

became more (0 0 2) oriented after annealing, as shown in Fig. 2. The annealing process seems to promote the regrowth of the original (0 0 2) oriented grains, along with the reduction of the (1 0 1) and (1 0 2) reflections. Therefore, the thermal annealing process supplies the system with the necessary energy for the rearrangement of the Al–N network. The regrowth of the original (0 0 2) grains can be attributed to the crystallization of amorphous material or to the annihilation of structural defects, which previously allowed the coalescence of grains. A different behavior was observed for the type II films. The XRD patterns reveal that the structure of the as deposited samples is maintained after the annealing at 900 ◦ C, the only noticeable effect being an increase in the grain size of the original (0 0 2) oriented crystallites. The main reason of this different structural behavior could be attributed to the low crystallinity of the as-deposited coating, which probably would require a higher annealing temperature for the recrystallization process to occur. 3.2. XPS analysis The chemical state and distribution of the elements at the surface level of the AlN coatings was assessed by XPS, in order to try to understand the possible interactions that might occur between elements, before and after subjecting the samples to the thermal treatment at 900 ◦ C. Table 3 summarizes the chemical compositions of the as-deposited and annealed AlN films, along with the sputtering conditions. The presence of O and C content in the film can be attributed to the residual gas in the sputtering chamber. From Table 3, it is observed that the nitrogen pressure has an impact on the content of Al, N and O in the AlN thin film. As the reactive gas pressure decreases, both the contents of Al and N gradually increase, while the oxygen content reduces: the aluminum content increases from 41.07% to 46.24% and the nitrogen increases from 26.43% to 33.68%, while the oxygen reduces from 23.10% to 14.33%. The oxygen incorporated into the AlN film deposited at a nitrogen pressure of 0.05 mbar is present in a lower quantity due to the lower adhesion coefficient of oxygen at low pressures. The deposition time did not have a significant influence on the chemical composition of the coatings. The annealing process did not induce an increase in O content because the thermal treatments where conducted in vacuum. The chemical structure of the as-deposited and annealed AlN films was determined by examining the Al 2p and N 1s corelevel spectrum, in order to determine the influence of thermal treatments on the chemical bonding. The deconvolution of Al 2p spectrums of the as-deposited AlN films (Fig. 3a) indicates that Al forms separate bonds with oxygen and nitrogen. The binding energy (BE) of Al N in the Al 2p spectrum is found in the range of 73.30–73.38 eV, confirming the formation of AlN [26], while the Al O bond is located at a higher BE in the range of 74.36–74.53 eV. The position of the Al O bond matches with the one found in Al2 O3 [26], although it is centered at a BE value lower than that reported for bulk Al2 O3 (∼75.5 eV) [27]. In Fig. 3c one can observe that the area under the Al O sub-peak is large,

Please cite this article in press as: M. P˘atru, et al., Structural, mechanical and piezoelectric properties of polycrystalline AlN films sputtered on titanium bottom electrodes, Appl. Surf. Sci. (2015), http://dx.doi.org/10.1016/j.apsusc.2015.07.022

G Model APSUSC-30747; No. of Pages 12

ARTICLE IN PRESS M. P˘atru et al. / Applied Surface Science xxx (2015) xxx–xxx

5

Fig. 3. XPS spectra of the as grown AlN samples concerning the a. Al 2p; b. N 1s.

suggesting that the oxygen incorporated during the sputtering process is preferentially bonded to the Al atoms. From a thermodynamics point of view, Al prefers to react with O2 when the contents of N2 and O2 are at the same level at room temperature [28]. Therefore, the sputtered Al atoms will react with the

residual oxygen from the sputtering chamber, which causes the formation of alumina. The high resolution XPS spectrum of the N 1s peaks (Fig. 3b) indicates the formation of AlN (at a BE in the range of 397.30–397.56 eV) and aluminum oxynitride (NO/Al at a BE in the range of 398.41 – 398.64 eV). Among them, it seems

Please cite this article in press as: M. P˘atru, et al., Structural, mechanical and piezoelectric properties of polycrystalline AlN films sputtered on titanium bottom electrodes, Appl. Surf. Sci. (2015), http://dx.doi.org/10.1016/j.apsusc.2015.07.022

G Model APSUSC-30747; No. of Pages 12 6

ARTICLE IN PRESS M. P˘atru et al. / Applied Surface Science xxx (2015) xxx–xxx

Fig. 4. XPS spectra of the annealed at 900 ◦ C AlN samples concerning the a. Al 2p; b. N 1s.

that the Al O N bond is the strongest, indicating that the surface of the AlN coatings has absorbed an important amount of N2 molecules. Since the XPS is only probing the surface of the thin films (in this specific case, the first ∼10–20 nm) and taking

into account the XRD analysis results, it is assumed that the aluminum oxynitride layer would be amorphous (there are no visible diffraction peaks that could be attributed to AlON, which would suggest a crystalline structure) and with a very low thickness.

Please cite this article in press as: M. P˘atru, et al., Structural, mechanical and piezoelectric properties of polycrystalline AlN films sputtered on titanium bottom electrodes, Appl. Surf. Sci. (2015), http://dx.doi.org/10.1016/j.apsusc.2015.07.022

G Model APSUSC-30747; No. of Pages 12

ARTICLE IN PRESS M. P˘atru et al. / Applied Surface Science xxx (2015) xxx–xxx

7

Fig. 5. SEM analysis – series no.1 (thickness of the AlN film ∼2.1 ␮m).

Fig. 6. SEM analysis – series no.2 (thickness of the AlN film ∼2.4 ␮m).

Fig. 7. SEM analysis – series no.3 (thickness of the AlN film ∼4.3 ␮m).

Furthermore, Al2 O3 could be introduced into the AlN structure as various forms of defects [27]. For the AlN films annealed at 900 ◦ C the main change, compared to the as-deposited samples, is inferred from the intensity of the two peaks associated with the Al N bonding (in the Al 2p, but especially in the N 1s spectrum, as seen in Fig. 4). This increase in Al N bonding during the annealing process indicates the diffusion of nitrogen into the film. It is also assumed that the duration of the annealing process is sufficient for the atoms to acquire enough kinetic energy and occupy relative equilibrium positions, which results in a better crystalline film, as demonstrated in the previous section (XRD analysis). Furthermore, the annealing process minimizes the structural defects and forms a better stoichiometric material. One observation needs to be mentioned for the samples sputtered with a nitrogen pressure of 0.1 mbar for 2 h (Fig. 4a.1). The annealing process causes in these samples the presence of Al OH bonding located at a binding energy of 77.38 eV, which is attributed

to the formation of aluminum hydroxide Al (OH)3 according to the Handbook of Monochromatic XPS Spectra [26]. It is in fact commonly established that alumina easily hydrates under ambient moisture exposure. The formation of hydroxides in the surface of AlN coatings depends on the permeation characteristics of the films which are closely related to the film morphology and surface defects.

3.3. SEM analysis If AlN films should be used as piezoelectric materials, the microstructure and surface morphology will have an important impact on the device performance. As the surface acoustic waves (SAW) only propagate on the surface, all the energy is confined almost within a wavelength from the surface to the inside [29]. On the other hand, a non-homogeneous microstructure, e.g., voids and dislocation cracks, would not only affect the SAW velocity and

Please cite this article in press as: M. P˘atru, et al., Structural, mechanical and piezoelectric properties of polycrystalline AlN films sputtered on titanium bottom electrodes, Appl. Surf. Sci. (2015), http://dx.doi.org/10.1016/j.apsusc.2015.07.022

G Model APSUSC-30747; No. of Pages 12 8

ARTICLE IN PRESS M. P˘atru et al. / Applied Surface Science xxx (2015) xxx–xxx

Fig. 8. AFM representative images of the AlN films deposited under different nitrogen flow pressures and deposition times: (a) 0.1 mbar – 2 h; (b) 0.05 mbar – 2 h; (c) 0.1 mbar – 4 h.

electromechanical coefficient, but also will result in transmission loss. The surfaces of the coatings, observed under SEM at 800× magnification, for the films deposited using high nitrogen pressure, exhibit a significant number of deposition features like porosities, uncovered areas or growth defects (Figs. 5a and 7a). These features are enhanced as a result of the annealing process at 900 ◦ C, one observing the appearance of cracks (Figs. 5d and 7d), along with the physical absorption of H2 O as demonstrated by the XPS analysis. Using lower nitrogen pressures, along with an optimum deposition rate, an AlN coating with the lowest level of surface defects was obtained, without cracks after annealing (Fig. 6a and d). The

SEM examination of the surface at 50000× magnification reveals a change in the AlN film morphology depending on the nitrogen pressure. A compact and homogenous surface morphology is observed for coating no. 2, which was deposited with 0.05 mbar nitrogen pressure (Fig. 6b). In contrast, the sample deposited with 0.1 mbar nitrogen pressure exhibits on the surface crystallites which begin to aggregate in grains (coating no. 1 – Fig. 5b). This aggregation becomes well defined when the sputtering time is doubled (coating no.3 – Fig. 7b). The cross-section SEM examination (Figs. 5c, 6c, and 7c) shows that all the samples have the same architecture: an AlN piezoelectric layer with different thicknesses for different samples, sputtered

Please cite this article in press as: M. P˘atru, et al., Structural, mechanical and piezoelectric properties of polycrystalline AlN films sputtered on titanium bottom electrodes, Appl. Surf. Sci. (2015), http://dx.doi.org/10.1016/j.apsusc.2015.07.022

G Model APSUSC-30747; No. of Pages 12

ARTICLE IN PRESS M. P˘atru et al. / Applied Surface Science xxx (2015) xxx–xxx

3.5. Hardness and elastic modulus The AlN/Ti coatings have been subjected to investigations concerning their respective mechanical properties, in order to understand how the microstructure of the deposited material relates to the mechanical behavior of the film/substrate system. The hardness and elastic modulus should be good indicators for the practical use of these films, especially for the proposed application. Furthermore, the properties of these devices might be seriously impaired due to the micro loading during machining, packaging and utilizing. The method proposed by Oliver and Pharr [31] was used to extract the mechanical properties from the load–displacement nanoindentation curve. The measured load–displacement curve is represented in Fig. 9a, while the measured hardness and elastic modulus as function of the penetration depth are presented in Fig. 9b and c, respectively. The data is obtained by measuring the reaction force as function of the penetration depth, as the indenter tip is pushed into the film under tightly controlled conditions. The force required to indent to a given depth provides a measurement of the hardness of the film, while the recovery of the indenter indicates the elasticity. The thickness of the deposited AlN thin films varies between 1.5 and 4.3 ␮m, while the maximum indentation depth was set at 2000 nm. The average values for hardness and elastic modulus were determined in various intervals of film thickness, thus obtaining their variation as function of the penetration depth, from the surface toward the Ti layer. The measurement data obtained in the 0–200 nm interval of penetration depths was not taken into consideration, as at very low loads inconclusive hardness values are recorded as a consequence of the indentation size effect (ISE). The ISE is often manifested as an increase in hardness with decreasing indentation depth for indenters such as pyramids and it is caused in general by the surface roughness, the presence of hard oxides on the sample surface or by indenter tip blunting, rather than by a true material effect [32]. Table 4 presents the values of hardness and Young’s modulus for AlN films as reported in the literature, compared with those measured on the samples presented in this work. As a consequence of

Young's modulus [GPa]

Atomic force microscopy was used to evaluate the film surface morphology and roughness. The AFM micrographs confirm the polycrystalline, columnar grain structure (Fig. 8). The average root mean square values of roughness for the deposited films are in the range of 7.2–25.3 nm over a 5 ␮m × 5 ␮m surface area. The roughness of the AlN coatings increases for the larger thickness AlN films, and also for the higher nitrogen pressure samples. In general, the surface roughness of a piezoelectric film is required to be less than 30 nm when it is used for SAW devices [30]. From this point of view, the surface roughness of the analyzed coatings is suited for SAW applications. In polycrystalline films, a direct correlation between the surface morphology, roughness and crystallinity is difficult to be established. As it has been observed after the SEM examination of the AlN films, the nanometric crystallites aggregate in grains. This aggregation process depends on several factors such as crystallite geometry and size, chemical surface composition, energy of the crystallites, in turn dependent on the deposition conditions (nitrogen pressure, time of sputtering, etc.).

Coating no.2 - N2 pressure=0.05 mbar

Coating no.3 - N2 pressure=0.1 mbar

c. 280.9 261.1 251.2 245.3 242.5 237.7 236.3 234.0

450 251.9 252.7 254.9 253.3 248.3 244.0 245.0 242.2

300

150

40

Hardness [GPa]

3.4. AFM analysis

Coating no.1 - N2 pressure=0.1 mbar

229.0 230.6 234.0 232.1 231.9 231.8 234.9 231.8

b.

30

16.85 14.88 13.59 12.47 11.75 11.35 11.00 10.70

20

10.73 11.59 12.26 12.07 11.80 11.47 11.18 10.90

10 10.82 11.36 10.96 10.62 10.35 10.06

9.81

9.75

800 0

Load on sample [mN]

on a Ti thin film that has the role of bottom electrode for the whole sensor coating. The cross-section micrographs were used to estimate the film thicknesses. It can be observed that the columnar growth mode of the sputtered AlN coatings is enhanced when the film thickness is larger.

9

a.

600 400 200 0 -200

0

200

400

600

800 1000 1200 1400 1600 1800 2000 2200

Displacement into surface [nm] Fig. 9. Indentation test on as-deposited AlN films: (a) load on sample vs. displacement into surface; (b) hardness vs. displacement into surface; (c) Young’s modulus vs. displacement into surface. Table 4 Mechanical properties of AlN films investigated in this study and those reported in the literatures. Literature

AlN coatings

H [GPa]

E [GPa]

Barshilia et al. [35] Mortet et al. [33] Jian et al. [34] This work

(0 0 2)-oriented

12

225

(0 0 2)-oriented

22

300

(1 0 3)-oriented

14.3–15.3–21.4

115.4–155.7–215.3

Polycrystalline oriented

10.82–16.85–10.73 229.0–280.9–251.9

the structure anisotropy of the AlN films, the hardness and Young’s modulus could vary as function of the crystal orientation and consequently with the deposition conditions. Endorsing this observation, Mortet et al. [31,33] report a value for hardness twice higher than the bulk value for aluminum nitride: 22 GPa measured on an AlN film with (0 0 2) orientation. Meanwhile, the hardness and Young’s modulus values of (1 0 3) oriented AlN film are reported to be 21.4 and 215 GPa, respectively [32,34]. In this work the near surface hardness of the studied AlN films is 10.82 for sample 1, 16.85 for sample 2 and 10.7 GPa for sample 3, respectively, while Young’s modulus is 229.0, 280.9 and 251.9 GPa, respectively. The surface hardness was obtained by averaging measurement data obtained in the interval 200–400 nm of film thickness, in order to avoid the indentation size effect. With the increase of indentation depth, the coating hardness gradually decreases until it reaches a value of ∼10 GPa (Fig. 9). The main contributors to this hardness decrease are the poorer substrate mechanical characteristics. Comparing the

Please cite this article in press as: M. P˘atru, et al., Structural, mechanical and piezoelectric properties of polycrystalline AlN films sputtered on titanium bottom electrodes, Appl. Surf. Sci. (2015), http://dx.doi.org/10.1016/j.apsusc.2015.07.022

G Model

ARTICLE IN PRESS

APSUSC-30747; No. of Pages 12

M. P˘atru et al. / Applied Surface Science xxx (2015) xxx–xxx

10 Coating no.1 - N2 pressure=0.1 mbar

Coating no.2 - N2 pressure=0.05 mbar

Young's modulus [GPa]

Coating no.3 - N2 pressure=0.1 mbar

c. 280.6 240.9 229.3 222.8 218.2 217.4 213.1

450

298.0 274.7 259.1 241.9 234.3 223.4 221.3 220.8

300

150

Hardness [GPa]

40

279.1 245.1 227.7 221.5 213.0 209.7 205.1

b.

30

7.41

19.35 11.79 8.79

16.23 14.59 13.72 11.42

6.47

5.93

5.57

9.77

8.81

8.04

Table 5 Piezoelectric coefficients measured on samples as deposited and after annealing at various temperature T = 300 ◦ C, 600 ◦ C and 900 ◦ C. Sputtering conditions Coating no◦

d33 [pC/N] as deposited

d33 [pC/N] T = 300o C

d33 [pC/N] T = 600o C

d33 [pC/N] T = 900o C

No. 1 Nitrogen flow: 0.1 mbar Sputtering time: 2 h No. 2 Nitrogen flow: 0.05 mbar Sputtering time: 2 h No. 3 Nitrogen flow: 0.1 mbar Sputtering time: 4 h

2.5 ± 0.3

2.5 ± 0.3

2.4 ± 0.3

4.0 ± 0.3

3.2 ± 0.2

2.7 ± 0.2

2.7 ± 0.3

4.3 ± 0.3

2.7 ± 0.3

2.8 ± 0.2

2.6 ± 0.3

3.6 ± 0.3

7.50

20

3.6. Piezoelectric properties 10 16.49 12.55 9.27

7.79

6.85

6.23

5.89

Load on sample [mN]

800 0 600

a.

400 200 0 -200

0

200

400

600

800 1000 1200 1400 1600 1800 2000 2200

Displacement into surface [nm] Fig. 10. Indentation test on AlN films after post-annealing at 900 ◦ C: (a) load on sample vs. displacement into surface; (b) hardness vs. displacement into surface; (c) Young’s modulus vs. displacement into surface.

mechanical properties of the three coatings, it can be observed that the surface hardness increases from ∼11 to ∼17 GPa when the nitrogen pressure decreases from 0.1 to 0.05 mbar. After the post-deposition annealing treatment at 900 ◦ C in vacuum, the film hardness increases up to 16.49 GPa for sample 1, 19.35 GPa for sample 2 and 16.23 GPa for sample 3 respectively, followed by a sharp decrease with the indentation depth until it reaches a value of 6–8 GPa (Fig. 10), similar to the substrate hardness after annealing at 900 ◦ C (∼6 GPa). The mechanical behavior after annealing can be explained through the fact that in polycrystalline AlN films the grain boundaries may act as channels for nitrogen incorporation inside the coating, which promote at high temperatures the diffusion of N into the films, as it was mentioned in the XPS analysis section, causing a surface hardening of the film.

In order to test the stability of AlN coatings at higher temperatures, the piezoelectric response was measured at room temperature on the as-deposited samples and after post-deposition annealing treatments performed at 300 ◦ C, 600 ◦ C and 900 ◦ C (Table 5). The post-deposition heat treatments were found to cause an interesting effect on the structural and piezoelectric properties of all the series of coatings considered in this paper. The piezoelectric tests results show an optimum value for the piezoelectric coefficient, in the range of room temperature to 600 ◦ C, with a slight decrease of the piezoelectric response for higher annealing temperatures. After the 900 ◦ C annealing an improvement of the piezoelectric properties of the highly (10×) oriented AlN films was noticed. This particular film exhibited after the annealing process a shift toward c-axis (0 0 2) orientation. A possible explanation for this phenomenon is related to the fact that the annealing process minimizes the structural defects and forms a better stoichiometric material (as observed from the XPS analysis) with an improved crystal quality (as observed from the XRD analysis). This results in an overall increase of the piezoelectric properties. These results are encouraging considering that in previous investigations AlN polycrystalline films with no preferred orientation did not undergo significant changes in their structural properties or in their piezoelectric response. Vergara et al. [36] obtained an improvement in the crystal quality after annealing only for highly (0 0 2) oriented AlN films, but not accompanied by a significant improvement of their piezoelectric properties. Taking into account the results presented herein, it can be stated that the samples deposited using 0.05 mbar nitrogen pressure and at a sputtering time of 2 h, exhibit improved thermal stability concerning their structural, mechanical and piezoelectric properties, when compared to the other samples presented in this paper. For these samples, Ti top electrodes were deposited on the AlN films (as

Fig. 11. EDX mapping (element distribution images).

Please cite this article in press as: M. P˘atru, et al., Structural, mechanical and piezoelectric properties of polycrystalline AlN films sputtered on titanium bottom electrodes, Appl. Surf. Sci. (2015), http://dx.doi.org/10.1016/j.apsusc.2015.07.022

G Model APSUSC-30747; No. of Pages 12

ARTICLE IN PRESS M. P˘atru et al. / Applied Surface Science xxx (2015) xxx–xxx

shown in the EDX mapping analysis in Fig. 11) in order to obtain an acoustic resonator structure (Ti electrode/AlN piezoelectric film/Ti electrode). For this particular configuration, a high piezoelectric response of ∼12pC/N was measured, indicating that the Ti/AlN/Ti multilayered structure has significant potential to be embedded into wear resistant coatings as a piezoelectric sensor, in order to monitor the wear rate and the remaining coating life. The piezoelectric constants measured on ZnO and quartz, which are used in various industries, are 2–13pC/N and 1.4–1.9pC/N, respectively [35]. All things considered, reactively sputtered aluminum nitride thin films, using certain deposition conditions, such as the ones presented in this paper, are promising candidates for piezoelectric wear monitoring sensors. 4. Conclusions The main purpose of this work was to present and to correlate the change of structural, mechanical and piezoelectric properties, measured on AlN coatings, with the processing parameters and post-deposition heat treatments at various temperatures. The aluminum nitride films were magnetron sputtered on Ti films, which in turn were deposited on 100Cr6 steel substrates. Three sets of AlN films were prepared by varying the reactive gas pressure and the deposition time. The AlN films exhibited significant differences, firstly related to the deposition parameters, for the as-deposited samples, and secondly, related to the annealing temperature. An optimized AlN/Ti coating system has been identified from the studied samples. For this particular coating, the following deposition parameters were used: a nitrogen pressure of 0.05 mbar and a deposition time of 2 h. The deposition at a lower nitrogen pressure seems to prevent the contamination of the film with oxygen and promotes a densification of the microstructure, an improvement of the crystal quality and good mechanical properties. Good piezoelectric properties were also achieved for these coatings. The obtained results are consistent with those reported in the literature [20,25] and indicate the formation of highly (1 0 1), (1 0 2) and (1 0 3) oriented AlN films with piezoelectric and mechanical properties suitable for practical implementation. The vacuum thermal annealing process at high temperature led to an improvement in the crystal quality of the AlN films that was accompanied by a clear increase of the piezoelectric response. The mechanical behavior after annealing shows as well a significant increase of the surface hardness and Young’s modulus values, followed by a sharp decrease with the indentation depth. Thus, the thermal annealing energy has promoted not only the rearrangement of Al–N network, but also the occurrence of a nitriding process of unsaturated Al atoms, causing a superficial hardening of the film. Future investigations are needed for establishing the optimal annealing temperature, in order to reach the best compromise between mechanical and piezoelectric properties of these coatings. Acknowledgments This work was supported by FEDER through the COMPETE Program and by the Portuguese Foundation for Science and Technology (FCT) in the framework of the Strategic Project PEST-C/FIS/UI607/2014 and the project Matepro – Optimizing Materials and Processes”, ref. NORTE-07-0124-FEDER-000037”, co-funded by the “Programa Operacional Regional do Norte” (ON.2 – O Novo Norte), under the “Quadro de Referência Estratégico Nacional” (QREN), through the “Fundo Europeu de Desenvolvimento Regional” (FEDER). P. M. acknowledges support from FCT (SFRH/BPD/96227/2013 and SFRH/BD/88397/2012 grants respectively).

11

References [1] S. Joblot, E. Feltin, E. Beraudo, P. Vennéguès, M. Leroux, F. Omnès, M. Laügt, Y. Cordier, Hexagonal c-axis GaN layers grown by metal organic vapor-phase epitaxy on Si (0 0 1), J. Cryst. Growth 280 (2005) 44–53. [2] C. Duquenne, M.P. Besland, P.Y. Tessier, E. Gautron, Y. Scudeller, D. Averty, Thermal conductivity of aluminum nitride thin films prepared by reactive magnetron sputtering, J. Phys. D 45 (2012) 015301. [3] V. Mortet, K. Haenen, J. Potmesil, M. Vanecek, M. D’Olieslaeger, Diamond – application to piezoelectric bimorph cantilever sensors, Phys. Status Solidi A 203 (2006) 3185–3190. [4] V. Brien, P. Pigeat, Correlation between the oxygen content and the morphology of AlN films grown by r.f. magnetron sputtering, J. Cryst. Growth 310 (2008) 3890–3895. [5] Z. Vashaei, T. Aikawa, M. Ohtsuka, H. Kobatake, H. Fukuyama, S. Ikeda, K. Takada, Influence of sputtering parameters on the crystallinity and crystal orientation of AlN layers deposited by RF sputtering using the AlN target, J. Cryst. Growth 311 (2009) 459–462. [6] T. Kumada, M. Ohtsuka, K. Takada, H. Fukuyama, Influence of sputtering conditions on crystalline quality of AlN layers deposited by RF reactive sputtering, Phys. Status Solidi C 8 (2011) 1520–1523. [7] T. Kumada, M. Ohtsuka, K. Takada, H. Fukuyama, Influence of sputter power and N2 gas flow ratio on crystalline quality of AlN layers deposited at 823K by reactive sputtering, Phys. Status Solidi C 9 (2012) 515–518. [8] M. Ihihara, S.J. Li, H. Yumoto, K. Akashi, Y. Ide, Control of preferential orientation of AlN films prepared by the reactive sputtering method, Thin Solid Films 316 (1998) 152. [9] X.H. Xu, H.S. Wu, C.J. Zhang, Z.H. Jin, Morphological properties of AlN piezoelectric thin films deposited by DC reactive magnetron sputtering, Thin Solid Films 388 (2001) 62. [10] D.Y. Wang, Y. Nagahata, M. Masuda, Y. Hayashi, Effect of non-stoichiometry upon optical properties of radio frequency sputtered Al–N thin film formed at various sputtering pressures, J. Vac. Sci. Technol. A 14 (1996) 3092. [11] R.N. Tait, A. Mirfazli, Low temperature aluminum nitride deposition on aluminum by RF reactive sputtering, J. Vac. Sci. Technol. A 19 (2001) 1586. [12] L. Wu, S. Wu, H.T. Song, Influence of sputtering pressure on physical structure of AlN thin films prepared on Y-128◦ LiNbO3 by RF magnetron sputtering, J. Vac. Sci. Technol. A 19 (2001) 167. [13] K. Sugiyama, K. Taniguchi, K. Kuwabara, Preparation of oriented aluminum nitride films by radio frequency reactive sputtering, J. Mater. Sci. Lett. 9 (1990) 489. [14] C.C. Cheng, Y.C. Chen, R.C. Horng, H.J. Wand, W.R. Chen, E.K. Lari, Growth of c-axis oriented aluminum nitride films on GaAs substrates by reactive RF magnetron sputtering, J. Vac. Sci. Technol. A 16 (1998) 3335. [15] M. Clement, E. Iborra, J. Sangrador, A. Sanz-Herva, L. Vergara, M. Aguilar, Influence of sputtering mechanisms on the preferred orientation of aluminum nitride thin films, J. Appl. Phys. 94 (2003) 1495. [16] X.H. Xu, H.S. Wu, C.J. Zhang, Z.H. Jin, Morphological properties of AlN piezoelectric thin films deposited by DC reactive magnetron sputtering, Thin Solid Films 388 (2001) 62–67. [17] C.C. Cheng, Y.C. Chen, H.J. Wang, W.R. Chen, J. Vac. Sci. Technol. A14 (1996) 2238–2242. [18] H. Okano, Y. Takahashi, T. Tanaka, K. Shibata, S. Nakano, Preparation of c-axis oriented AlN thin films by low-temperature reactive sputtering, Jpn. J. Appl. Phys. 31 (1992) 3446–3451. [19] J.C. Doll, B.C. Petzold, B. Ninan, R. Mullapudi, B.L. Pruitt, Aluminum nitride on titanium for CMOS compatible piezoelectric transducers, J. Micromech. Microeng. 20 (2) (2010) 25008. [20] A. Ababneh, M. Alsumady, H. Seidel, T. Manzaneque, J. Hernando-García, J.L. Sánchez-Rojas, A. Bittner, U. Schmid, c-Axis orientation and piezoelectric coefficients of AlN thin films sputter deposited on titanium bottom electrodes, Appl. Surf. Sci. 259 (2012) 59–65. [21] P.S. Henderson, P.J. Kelly, R.D. Amell, H. Backer, J.W. Bradley, Investigation into the properties of titanium based films deposited using pulsed magnetron sputtering, Surf. Coat. Technol. 174 (2003) 779–783. [22] B.D. Cullity, S.R. Stock, Elements of X-ray Diffraction, Prentice-Hall, Englewood Cliffs, NJ, 2001. [23] M. Akiyama, H.R. Kokabi, K. Nonaka, K. Shobu, T. Watanabe, Influence of substrate temperature on physical structure of AlN thin films prepared on polycrystalline MoSi2 by rf magnetron sputtering, J. Am. Ceram. Soc. 78 (12) (1995) 3304–3308. [24] A.T. Trana, O. Wunnickeb, G. Pandrauda, M.D. Nguyenc, H. Schellevisa, P.M. Sarroa, Slender piezoelectric cantilevers of high quality AlN layers sputtered on Ti thin film for MEMS actuators, Sens. Actuators A: Phys. 1 (2013, November) 118–123. [25] T. Kamohara, M. Akiyama, N. Ueno, K. Nonaka, N. Kuwano, Influence of aluminum nitride interlayers on crystal orientation and piezoelectric property of aluminum nitride thin films prepared on titanium electrodes, Thin Solid Films 515 (2007) 4565–4569. [26] B.V. Crist, Handbooks of Monochromatic XPS Spectra, XPS International LLC, Mountain View, CA, USA, 2005. [27] P. Motamedi, K. Cadien, XPS analysis of AlN thin films deposited by plasma enhanced atomic layer deposition, Appl. Surf. Sci. 315 (2014) 104–109.

Please cite this article in press as: M. P˘atru, et al., Structural, mechanical and piezoelectric properties of polycrystalline AlN films sputtered on titanium bottom electrodes, Appl. Surf. Sci. (2015), http://dx.doi.org/10.1016/j.apsusc.2015.07.022

G Model APSUSC-30747; No. of Pages 12 12

ARTICLE IN PRESS M. P˘atru et al. / Applied Surface Science xxx (2015) xxx–xxx

[28] H.C. Lee, G.H. Kim, S.K. Hong, Y.J. Yong, C.H. Chun, K.Y. Lee, J.Y. Lee, Influence of sputtering pressure on the microstructure evolution of AlN thin films prepared by reactive sputtering, Thin Solid Films 261 (1995) 148–153. [29] J.J.B. EunHyun, O. MinSuk, H. JunHee, J. JinWoo, Industrial applications of cathodic-arc and RF/DC magnetron plasma, IEEE Int. Conf. Plasma Sci. 163 (1995). [30] D. Yang, R. Jonnalagadda, B.R. Rogers, J.T. Hillman, R.F. Foster, T.S. Cale, Texture and surface roughness of PRCVD aluminum films, Thin Solid Films 332 (1998) 312–318. [31] W.C. Oliver, G.M. Pharr, An improved technique for determining hardness and elastic modulus using load and displacement sensing indentation experiments, J. Mater. Res. 7 (6) (1992) 1564–1583.

[32] T.F. Page, W.C. Oliver, C.J. McHargue, J. Mater. Res. 7 (1992) 450. [33] V. Mortet, M. Nesladek, K. Haenen, A. Morel, M. D’Olieslaeger, M. Vaneceka, Physical properties of polycrystalline aluminum nitride films deposited by magnetron sputtering, Diamond Relat. Mater. 13 (2004) 1120–1124. [34] S.R. Jian, G.J. Chena, J.S.C. Jangb, Y.S. Lai, Nanomechanical properties of AlN(1 0 3) thin films by nanoindentation, J. Alloys Compd. 494 (2010) 219–222. [35] H.C. Barshilia, B. Deepthi, K.S. Rajam, Growth and characterization of aluminum nitride coatings prepared by pulsed-direct current reactive unbalanced magnetron sputtering, Thin Solid Films 516 (2008) 4174. [36] L. Vergara, J. Olivares, E. Iborra, M. Clement, A. Sanz-Hervás, J. Sangrador, Effect of rapid thermal annealing on the crystal quality and the piezoelectric response of polycrystalline AlN films, Thin Solid Films 515 (2006) 1814–1818.

Please cite this article in press as: M. P˘atru, et al., Structural, mechanical and piezoelectric properties of polycrystalline AlN films sputtered on titanium bottom electrodes, Appl. Surf. Sci. (2015), http://dx.doi.org/10.1016/j.apsusc.2015.07.022